Comparative Analysis of RED Queue Variants for Data Traffic Reduction Over Wireless Network

Wide usage of internet and connectivity of more devices on the wireless network increases the usage of TCP layer that increases heavily leading to congestion. Congestion in the network is responsible for heavy traffic, increasing delay, and packet loss, which reduces the quality of service. Various techniques are developed for congestion control in dense traffic networks. Random early detection (RED) is one of the widespread techniques for congestion control. In this article, the comparison of various techniques has been carried out and different parameters for the congestion control are also analyzed. The early congestion control (ECC) is employed, where the TCP header window filed is continuously by changing the congestion. In the three sections, random early detection (TRED) where dropping probability is calculated according to that decision is carried out for the congestion control, here connection setup is divided into three sections such method is recognized as three sections random early detection (TRED). In addition to these methods, other methods are suggested those are based on non-congestion notification that sends the notification about the congestion, fuzzy logic dimensions, and characterization of problems for the congestions in the RED, nonlinear packet loss and using cloud-based model reduce the congestion under the Hemi-rise cloud model (CRED), congestion avoidance mechanisms to enhance by Learning Automata Like (LAL) philosophy, where the complete five-step algorithm is used in efficient LAL random early detection. In this paper, we studied various congestion control techniques and analyzed, we also provide suggestions to improve the congestion control mechanism.